Interface circuit and semiconductor integrated circuit

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to an interface circuit and a semiconductor integrated circuit that are used for performing signal transmission with a transmission line, as well as a method for adjusting termination resistance. The present invention can be used as a high speed interface circuit such as a USB driver.

2. Description of the Prior Art

Conventionally, interface circuits are disposed at every place in an internal circuit and an external circuit of a computer. For example, DDR2 (Double Data Rate 2) standard is adopted as an interface circuit for high speed data transmission in an SDRAM that is used as a main memory of a computer. In addition, USB (Universal Serial Bus) standard is used commonly as an interface circuit between a computer and a peripheral device.

In such an interface circuit, it is necessary to match its output impedance with characteristic impedance of a transmission line such as a reception circuit or a cable so as to perform correct transmission by preventing reflections and losses of an output signal. In addition, it is necessary to adjust rise time and fall time (slew rate) of a waveform of the output signal to fall within an appropriate range. If it is too short, it may cause noises. If it is too long, the waveform may be deteriorated.

In order to deal with this problem conventionally, a function for adjusting and controlling termination resistance in the interface circuit is provided inside a memory LSI (Large Scale Integrated circuit) that is a semiconductor integrated circuit. For example, an ODT (On Die Termination) circuit that is provided to a DDR2 standard memory can adjust its termination resistance to a value of 75 ohms or 150 ohms. In other words, the ODT circuit has many transistor pairs of p-type and n-type that are connected in parallel so as to form a terminating resistor. The number of transistors that are connected in parallel actually is adjusted by a control signal supplied to their gates, so that a resistance value of the transistors is controlled to be equal to a resistance value of an external reference resistor (see U.S. Pat. No. 7,193,431).

In addition, in order to maintain the impedance and the slew rate at constant values even if a manufacturing process or an operational environment changes, a method is proposed in which output impedance of a plurality of output MOSFET's connected in parallel is adjusted by a first controlling portion that selects the number of output MOSFET's to be turned on, while a slew rate is adjusted by a second controlling portion that adjusts a drive signal of the output MOSFET to be turned on (see Japanese unexamined patent publication No. 2004-327602).

However, the interface circuit or the driver circuit described above is usually disposed inside the LSI, so values of on-resistance of on-chip resistors or transistors vary substantially in accordance with process conditions in the manufacturing steps of the LSI or its operating temperature. Therefore, the structure described in U.S. Pat. No. 7,193,431 includes an additional reference resistor provided to the outside of the LSI, which causes a problem that the number of components as well as an area for mounting components are increased due to the external reference resistor.

In addition, the circuit described in Japanese unexamined patent publication No. 2004-327602 uses an impedance adjustment code and a slew rate adjustment code for performing adjustment of the output impedance or the slew rate, but it only discloses that an external resistor should be used as a method for generating the codes.

In this way, although it is proposed that adjustment of the terminating resistor and adjustment of the slew rate should be performed in the interface circuit conventionally, the adjustment needs an external reference resistor and cannot be performed as an internal function of the LSI.

Thus, there are remaining problems in the conventional interface circuit, which are increases of the number of components and the area for mounting components, deterioration of waveform due to increase of reactance when an external resistor is used, and others.

SUMMARY OF THE INVENTION

An object of the present invention is to provide an interface circuit and a semiconductor integrated circuit, in which the entire interface circuit can be structured inside the LSI, and variation of process conditions or operating temperature can be compensated even if it occurs during the manufacturing steps, so that substantial adjustment of the terminating resistor can be performed.

Another object of the present invention is to maintain slew rate substantially constant even if variation of process conditions or temperature occurs.

An interface circuit according to an embodiment of the present invention includes a driver circuit made up of a combination of a plurality of transistors, a calibration circuit for performing selection of on and off of one or more of the plurality of transistors for adjusting on-resistance of the plurality of transistors, and a terminating resistor that is connected between an output side of the driver circuit and the transmission line. One or more of the plurality of transistors are turned on based on an output of the calibration circuit, so that a combination resistance value of the on-resistance of the plurality of transistors and the terminating resistor matches characteristic impedance of the transmission line.

Furthermore, the driver circuit, the calibration circuit and the terminating resistor may be formed on one identical semiconductor integrated circuit. The calibration circuit may be adapted to detect process variation of a transistor and a resistor formed on the semiconductor integrated circuit and temperature variation thereof.

Thus, the calibration circuit detects process variation and temperature variation of the driver circuit, the terminating resistor and the like, so that they are compensated and the terminating resistor can be adjusted substantially.

In addition, the plurality of transistors of the driver circuit may include a commonly used transistor that is not selected by the calibration circuit, so that a combination resistance value of the commonly used transistor and the terminating resistor is set to be close to the characteristic impedance of the transmission line, and one or more of the plurality of transistors except the commonly used transistor may be selected by the calibration circuit so that the combination resistance value of them is adjusted to match the characteristic impedance of the transmission line.

Note that the transmission line in the present invention includes not only cables and wires but also various ones to be a load of the output of the interface circuit. For example, the transmission line may be a connector, a reception circuit of the other end, a transmission circuit, a transceiver circuit, and the like.

According to the present invention, the entire interface circuit can be formed inside the LSI, and even if variation occurs in the process conditions of the manufacturing steps or in operating temperature, it can be compensated so that the terminating resistor can be adjusted substantially.

According to the present invention, even if variation occurs in the process conditions or in temperature, a slew rate can be maintained substantially constant.

BRIEF DESCRIPTION OF THE DRAWINGS

<figref idrefs="DRAWINGS">FIG. 1</figref> is a diagram showing an example of a basic structure of a semiconductor integrated circuit according to an embodiment of the present invention.

<figref idrefs="DRAWINGS">FIG. 2</figref> is a circuit diagram showing a driving portion of a first embodiment.

<figref idrefs="DRAWINGS">FIG. 3</figref> is a circuit diagram showing a calibration portion.

<figref idrefs="DRAWINGS">FIG. 4</figref> is a diagram showing an example of a switch circuit.

<figref idrefs="DRAWINGS">FIG. 5</figref> is a timing chart of an example of an operation of the calibration portion.

<figref idrefs="DRAWINGS">FIG. 6</figref> is a circuit diagram of the driving portion of a second embodiment.

<figref idrefs="DRAWINGS">FIG. 7</figref> is a circuit diagram of the driving portion of a third embodiment.

<figref idrefs="DRAWINGS">FIG. 8</figref> is a circuit diagram of the calibration portion of a fourth embodiment.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The invention will now be described in detail with reference to the attached drawings.

<figref idrefs="DRAWINGS">FIG. 1</figref> is a diagram showing an example of a basic structure of a semiconductor integrated circuit SK according to an embodiment of the present invention.

In <figref idrefs="DRAWINGS">FIG. 1</figref>, the semiconductor integrated circuit SK includes a driver circuit <b>12</b> made up of a combination of a plurality of MOSFET's (transistors) <b>101</b><i>a</i>-<b>101</b><i>d</i>, a calibration circuit <b>14</b> for performing selection of on and off for adjusting on-resistance of the plurality of MOSFET's <b>101</b><i>a</i>-<b>101</b><i>d</i>, and an interface circuit <b>1</b> made up of a terminating resistor <b>13</b> connected between the output side of the driver circuit <b>12</b> and a terminal TS for connection with external circuit and the like, which are formed in a single package.

The calibration circuit <b>14</b> detects process variation and temperature variation of a MOSFET <b>102</b> and a resistor <b>103</b> formed on the same semiconductor integrated circuit SK so as to turn on the plurality of MOSFET's <b>101</b><i>b</i>-<b>101</b><i>d </i>selectively. Thus, a combined resistance value of on-resistance of the plurality of MOSFET's <b>101</b><i>a</i>-<b>101</b><i>d </i>and the terminating resistor <b>13</b> is adjusted to be a desired value. Note that a pre-buffer circuit <b>11</b> for driving the driver circuit <b>12</b> is provided before the driver circuit <b>12</b>.

Hereinafter, a case where the interface circuit <b>1</b> is an interface for a USB device will be described with reference to examples of various embodiments.

First Embodiment

<figref idrefs="DRAWINGS">FIG. 2</figref> is a circuit diagram showing a driving portion <b>3</b> of the interface circuit <b>1</b> according to a first embodiment of the present invention, <figref idrefs="DRAWINGS">FIG. 3</figref> is a circuit diagram showing a calibration portion <b>4</b> of the interface circuit <b>1</b>, <figref idrefs="DRAWINGS">FIG. 4</figref> is a diagram showing an example of a switch circuit, and <figref idrefs="DRAWINGS">FIG. 5</figref> is a timing chart showing an example of an operation of the calibration portion <b>4</b>.

The interface circuit <b>1</b> shown in <figref idrefs="DRAWINGS">FIG. 2</figref> indicates a part of the internal circuit formed on a single LSI (semiconductor integrated circuit) that is incorporated in a USB device, for example, and has a part of functions of a transmitter according to the USB standard. The interface circuit <b>1</b> performs signal transmission between the USB device and an external transmission line, for example, a cable, a connector, a receiver circuit or a transceiver circuit of the other end, or the like.

In <figref idrefs="DRAWINGS">FIGS. 2 and 3</figref>, the interface circuit <b>1</b> is made up of the pre-buffer circuit <b>11</b>, the driver circuit <b>12</b>, the terminating resistor <b>13</b>, the calibration circuit <b>14</b> and the like. The driving portion <b>3</b> is made up of the pre-buffer circuit <b>11</b>, the driver circuit <b>12</b> and the terminating resistor <b>13</b>.

The pre-buffer circuit <b>11</b> is provided before the driver circuit <b>12</b> and drives the driver circuit <b>12</b> based on an input signal (input data) S<b>1</b> having two digital values. The pre-buffer circuit <b>11</b> includes two pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b</i>. Each of the pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b </i>is made up of a P-channel MOSFET <b>111</b> or <b>113</b> and an N-channel MOSFET <b>112</b> or <b>114</b> that are connected in series.

When the input signal S<b>1</b> of the driver circuit <b>12</b> becomes “H” or “L”, the output of the pre-buffer <b>11</b><i>a </i>is turned on (“H”) or off (“L”), while the output of the pre-buffer <b>11</b><i>b </i>is turned off or on.

Although one P-channel MOSFET and one N-channel MOSFET are shown in each of the pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b </i>in <figref idrefs="DRAWINGS">FIG. 2</figref>, it is possible that each of them has the plurality of MOSFET's connected in parallel or in series.

The driver circuit <b>12</b> includes two driver circuits <b>12</b><i>a </i>and <b>12</b><i>b </i>of P side and N side. Each of the driver circuits <b>12</b><i>a </i>and <b>12</b><i>b </i>is made up of a combination of a plurality of MOSFET'S.

More specifically, the driver circuit <b>12</b><i>a </i>includes five P-channel MOSFET's <b>120</b>-<b>124</b> connected in parallel. Four MOSFET's <b>121</b>-<b>124</b> among them are controlled to be turned on and off by switches <b>131</b>-<b>134</b>. Each of sources of the MOSFET's <b>120</b>-<b>124</b> is connected to a power supply Vdd. As the power supply Vdd, a constant voltage power supply of 3.3 volts is used, for example. In addition, a gate of one MOSFET <b>120</b> is connected directly, and gates of other MOSFET's <b>121</b>-<b>124</b> are connected via the switches <b>131</b>-<b>134</b> to the output terminal of the pre-buffer <b>11</b><i>a. </i>

In addition, the driver circuit <b>12</b><i>b </i>has five N-channel MOSFET's <b>125</b>-<b>129</b> connected in parallel. Four MOSFET's <b>126</b>-<b>129</b> among them are controlled to be turned on and off by switches <b>136</b>-<b>139</b>. Each of sources of the MOSFET's <b>125</b>-<b>129</b> is connected to the ground. A gate of one MOSFET <b>125</b> is connected directly, while gates of other MOSFET's <b>126</b>-<b>129</b> are connected via the switches <b>136</b>-<b>139</b> to the output terminal of the pre-buffer <b>11</b><i>b. </i>

Although one MOSFET is shown as each of the MOSFET's <b>120</b>-<b>129</b> in <figref idrefs="DRAWINGS">FIG. 2</figref>, a plurality of MOSFET's connected in parallel are actually used for each of them for adjusting their on-resistance. For example, a to j of MOSFET's are used for each of them, like “a” MOSFET's for the MOSFET <b>120</b>, “b” MOSFET's for the MOSFET <b>121</b>, and “c” MOSFET's for the MOSFET <b>122</b>. Note that a to j are integers, which can be different numbers or the same number. However, it is preferable to form the state where the driver circuit <b>12</b><i>a </i>and the driver circuit <b>12</b><i>b </i>are symmetric with each other for easy control. Note that the MOSFET's are “transistors” in the present invention. It is true in the following description, too.

The switches <b>131</b>-<b>139</b> are controlled to be turned on and off by control signals D<b>1</b>-D<b>6</b> from the calibration circuit <b>14</b>. For example, if the control signals D<b>1</b>-D<b>6</b> are on (“H”), the corresponding switches <b>131</b>-<b>134</b> and <b>136</b>-<b>139</b> are turned on. As such switches <b>131</b>-<b>134</b> and <b>136</b>-<b>139</b>, ones having various structures can be used. For example, an analog switch <b>21</b> having a gate circuit made up of two MOSFET's <b>211</b> and <b>212</b> of P-channel and N-channel can be used as shown in <figref idrefs="DRAWINGS">FIG. 4</figref>.

An end of the terminating resistor <b>13</b> is connected to the output side of the driver circuit <b>12</b>, while the other end is connected directly or indirectly to an external transmission line. For example, the other end of the terminating resistor <b>13</b> is connected to a pin terminal of the LSI for output. Since characteristic impedance of the USB transmission line is 45 ohms, a resistance value of the terminating resistor <b>13</b> is set to 39 ohms in the present embodiment, so that the output impedance of the interface circuit <b>1</b> matches the characteristic impedance.

The on-resistance of the MOSFET's <b>120</b> and <b>125</b> that are connected without using a switch is set to 6 ohms. Since the terminating resistor <b>13</b> and the on-resistance of the MOSFET's <b>120</b> and <b>125</b> are connected in series, the output impedance of the interface circuit <b>1</b> becomes 45 ohms (=39 ohms+6 ohms) in the case where the switches <b>131</b>-<b>134</b> and <b>136</b>-<b>139</b> are all off. However, the terminating resistor <b>13</b> and the on-resistance of the MOSFET's <b>120</b> and <b>125</b> vary in accordance with process conditions of the LSI in the manufacturing steps or operating temperature.

Therefore, in the present embodiment, the variation due to manufacturing conditions of the LSI and the operating temperature is detected by the calibration circuit <b>14</b>, and the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> are turned on selectively in accordance with a result of the detection so that the on-resistance thereof is connected in parallel with the on-resistance of the MOSFET's <b>120</b> and <b>125</b>. Thus, a value of a combination resistance of the on-resistance of the MOSFET's <b>120</b>-<b>129</b> and the terminating resistor <b>13</b> becomes constant regardless of variation of the process conditions in the manufacturing process of the LSI or operating temperature.

In this way, in the present embodiment, impedance matching is achieved to a certain extent with the terminating resistor <b>13</b> and the MOSFET <b>120</b> or the MOSFET <b>125</b>. Then, the process variation and the temperature variation is detected by the calibration circuit <b>14</b>, and the switches <b>131</b>-<b>134</b> and <b>136</b>-<b>139</b> are turned on in accordance with a result of the detection. Thus, the on-resistance of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> are selectively connected so that the entire termination resistance is changed for fine adjustment. In this way, the interface circuit <b>1</b> including the terminating resistor <b>13</b> can be structured inside the LSI. Even if process variation or temperature variation occurs, it is possible to adjust the terminating resistor so as to compensate the variation so that substantial impedance matching can be performed.

Next, the calibration circuit <b>14</b> that constitutes the calibration portion <b>4</b> will be described.

The calibration circuit <b>14</b> is formed on the same LSI as the pre-buffer circuit <b>11</b>, the driver circuit <b>12</b> and the terminating resistor <b>13</b> described above. Therefore, MOSFET's and registers used for them are manufactured by the same process steps, and environment conditions such as operating temperature are also the same for them.

The calibration circuit <b>14</b> is provided with a constant current circuit (constant current source) that supplies constant current to the MOSFET's and resistors. It generates voltage (drop voltage) corresponding to an actual resistance value every now and then by supplying constant current to the transistor or the resistor. If the drop voltage level is detected, the process variation or the temperature variation can be detected.

More specifically, in <figref idrefs="DRAWINGS">FIG. 3</figref>, the calibration circuit <b>14</b> includes an NMOS calibration circuit <b>14</b><i>a</i>, a resistor calibration circuit <b>14</b><i>b</i>, a PMOS calibration circuit <b>14</b><i>c </i>and a latch circuit <b>61</b>.

The NMOS calibration circuit <b>14</b><i>a </i>supplies constant current to an N-channel MOSFET <b>32</b> in the on-state from a constant current source <b>31</b>, and a voltage VC<b>1</b> generated by the constant current is detected by a detection circuit made up of four resisters <b>33</b>-<b>36</b> and two comparators <b>37</b> and <b>38</b>. More specifically, the voltage VC<b>1</b> is compared with a voltage (reference voltage) Vref<b>1</b> obtained by dividing the power supply Vdd by two resisters <b>33</b> and <b>34</b> by using the comparator (comparing circuit) <b>37</b>. If the voltage VC<b>1</b> exceeds the voltage Vref<b>1</b>, the output of the comparator <b>37</b> becomes “H”. In addition, the comparator <b>38</b> compares the voltage VC<b>1</b> with the voltage Vref<b>2</b> that is obtained by dividing the power supply Vdd by two resistors <b>35</b> and <b>36</b>. If the voltage VC<b>1</b> exceeds the voltage Vref<b>2</b>, the output of the comparator <b>38</b> becomes “H”. The outputs are retained by the latch circuit <b>61</b>, and the signals in the retained state are delivered as data D<b>1</b> and D<b>2</b>.

Note that the state “H” or “L” of the signal retained by the latch circuit <b>61</b> is updated every time when a trigger signal ST is supplied. The timing when such a trigger signal ST is supplied should be, for example, a timing just before the signal is delivered from the interface circuit <b>1</b> to the transmission line and/or a timing just before each frame in data transmission.

In the example shown in <figref idrefs="DRAWINGS">FIG. 5</figref>, a relationship of “Vref<b>1</b>&gt;Vref <b>2</b>” is satisfied. If the voltage VC<b>1</b> exceeds the voltage Vref <b>2</b>, the data D<b>2</b> becomes “H”. If it exceeds the voltage Vref<b>1</b>, both the data D<b>1</b> and D<b>2</b> become “H”.

In addition, the resistor calibration circuit <b>14</b><i>b </i>and the PMOS calibration circuit <b>14</b><i>c </i>are also structured similarly to the NMOS calibration circuit <b>14</b><i>a </i>described above. The output states “H” or “L” of comparators <b>47</b>, <b>48</b>, <b>57</b> and <b>58</b> are retained by the latch circuit <b>61</b> and are delivered as data D<b>3</b>, D<b>4</b>, D<b>5</b> and D<b>6</b>, respectively.

Furthermore, as for the data D<b>1</b>-D<b>6</b>, it is possible to adopt a structure in which if the data D with respect to the higher reference voltage Vref is “H”, the data D with respect to the lower reference voltage Vref becomes “L”. In addition, it is possible to adopt a structure in which the data D<b>1</b>-D<b>6</b> are combined by an appropriate logic so as to generate data D for a new control, which is delivered to the switches <b>131</b>-<b>134</b> and <b>136</b>-<b>139</b>.

In addition, values of the voltage Vref<b>1</b>-Vref<b>6</b> and relationship among them should be set so that compensation of the terminating resistor with respect to the process variation or the temperature variation can be performed effectively as the entire interface circuit <b>1</b> and that it falls within the range of specification of the interface circuit <b>1</b> of the LSI as a result. For example, the values of the voltage Vref<b>1</b>-Vref<b>6</b>, values of on-resistance of the MOSFET's <b>120</b>-<b>129</b>, and the like should be decided by a simulation.

For example, as described above, supposing that the terminating resistor <b>13</b> is 39 ohms, on-resistance of the MOSFET's <b>120</b> and <b>125</b> is 6 ohms, an error portion and a variation portion thereof should be compensated by adjusting the on-resistance of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b>. In this case, for example, on-resistance of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> may be set to satisfy the relationship of power of “2” like 1:2:4:8.

Note that the error and the variation of the terminating resistor <b>13</b> are smaller than those of the on-resistance of the MOSFET's <b>120</b> and <b>125</b> and have a relatively stable resistance value. Therefore, it is easier for adjustment to set the resistance value of the terminating resistor <b>13</b> larger than the on-resistance value of the MOSFET's <b>120</b> and <b>125</b>.

In the interface circuit <b>1</b> of the first embodiment described above, the impedance matching can be realized by adjusting the terminating resistor even if the process variation or the temperature variation occurs. However, since the number of MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> to be turned on varies, gate capacitance that is a load viewed from the pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b </i>varies, so the slew rate may vary. Therefore, an example of an interface circuit <b>1</b>B that can maintain the slew rate substantially by suppressing variation thereof will be described as a second embodiment.

Second Embodiment

In the second embodiment, fundamental portions are the same as those of the first embodiment, so only different portions will be described.

<figref idrefs="DRAWINGS">FIG. 6</figref> is a circuit diagram showing a driving portion <b>3</b>B of an interface circuit <b>1</b>B according to the second embodiment of the present invention.

In <figref idrefs="DRAWINGS">FIG. 6</figref>, the driving portion <b>3</b>B is made up of a pre-buffer circuit <b>11</b>, a driver circuit <b>12</b>, a terminating resistor <b>13</b> and a slope switching circuit <b>15</b>. The slope switching circuit <b>15</b> has two slope switching circuits <b>15</b><i>a </i>and <b>15</b><i>b. </i>

As understood from comparison between <figref idrefs="DRAWINGS">FIG. 6</figref> and <figref idrefs="DRAWINGS">FIG. 2</figref>, the source of a MOSFET <b>112</b> of a pre-buffer <b>11</b><i>a </i>in the driving portion <b>3</b>B is connected to the ground not directly but indirectly via the slope switching circuit <b>15</b><i>a</i>. In addition, the source of a MOSFET <b>113</b> of a pre-buffer <b>11</b><i>b </i>is connected to the power supply Vdd not directly but indirectly via the slope switching circuit <b>15</b><i>b. </i>

In the slope switching circuit <b>15</b><i>a</i>, MOSFET <b>152</b> and MOSFET's <b>160</b>-<b>164</b> constitute a current mirror circuit with the MOSFET <b>152</b> as a reference transistor. In other words, a ratio of current (reference current) flowing in the MOSFET <b>152</b> to current flowing in each of the MOSFET's <b>160</b>-<b>164</b> is decided by a ratio (current mirror ratio) between sizes (element sizes W and L) of the transistors. In other words, current flowing in each of the MOSFET's <b>160</b>-<b>164</b> becomes a value obtained by multiplying current flowing in the MOSFET <b>152</b> as the reference transistor by the current mirror ratio.

In addition, since the constant current I<b>1</b> from the constant current source <b>151</b> flows in the MOSFET <b>152</b>, current having a value obtained by multiplying the current I<b>1</b> by each current mirror ratio flows in each of the MOSFET's <b>160</b>-<b>164</b>.

As for the slope switching circuit <b>15</b><i>b </i>too, the MOSFET <b>153</b> and the MOSFET's <b>165</b>-<b>169</b> constitute a current mirror circuit with the MOSFET <b>153</b> as a reference transistor in the same manner. Therefore, current flowing in each of the MOSFET's <b>165</b>-<b>169</b> has a value obtained by multiplying current flowing in the MOSFET <b>153</b> as the reference transistor by the current mirror ratio.

In addition, since constant current I<b>2</b> from the constant current source <b>154</b> flows in the MOSFET <b>153</b>, current having a value obtained by multiplying the current I<b>2</b> by each current mirror ratio flows in each of the MOSFET's <b>165</b>-<b>169</b>.

The MOSFET's <b>161</b>-<b>164</b> and <b>166</b>-<b>169</b> are controlled to be turned on and off by the switches <b>171</b>-<b>174</b> and <b>176</b>-<b>179</b>, respectively.

The switches <b>171</b>-<b>174</b> and <b>176</b>-<b>179</b> are controlled to be turned on and off by the control signals D<b>1</b>-D<b>6</b> from the calibration circuit <b>14</b> shown in <figref idrefs="DRAWINGS">FIG. 3</figref>. For example, when the control signals D<b>1</b>-D<b>6</b> are on (“H”), the corresponding switches <b>171</b>-<b>174</b> and <b>176</b>-<b>179</b> are turned on.

As understood from <figref idrefs="DRAWINGS">FIG. 6</figref>, for example, when the control signal D<b>1</b> is turned on, the MOSFET <b>126</b> and the MOSFET <b>166</b> are turned on in synchronization with each other. When the control signal D<b>2</b> is turned on, the MOSFET <b>127</b> and the MOSFET <b>167</b> are turned on in synchronization with each other. In addition, when the control signal D<b>3</b> is turned on, the MOSFET <b>128</b> and the MOSFET <b>168</b>, as well as the MOSFET <b>123</b> and the MOSFET <b>164</b> are turned on in synchronization with each other. In this way, the MOSFET's <b>161</b>-<b>164</b> and <b>166</b>-<b>169</b> are adapted to be turned on and off in synchronization with on and off of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b>.

Therefore, when the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> are turned on for impedance matching, the MOSFET's <b>161</b>-<b>164</b> and <b>166</b>-<b>169</b> are turned on for supplying drive current corresponding to a change of gate capacitance due to it. In other words, when the number of MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> to be driven increases, the number of MOSFET's <b>161</b>-<b>164</b> and <b>166</b>-<b>169</b> that supply the drive current is increased so as to increase the current.

Since current that flows in the MOSFET's <b>152</b> and <b>153</b> as reference transistors and a current mirror ratio of the MOSFET's <b>161</b>-<b>164</b> and <b>166</b>-<b>169</b> are adjusted, appropriate drive current corresponding to a change of the gate capacitance can be supplied. Thus, it is possible to suppress variations of the rise time Tr and the fall time Tf of the signal waveforms delivered from the driver circuits <b>12</b><i>a </i>and <b>12</b><i>b </i>and to maintain them substantially constant.

In this way, according to the interface circuit <b>1</b>B of the second embodiment, even if process variation or temperature variation occurs, it is possible to realize impedance matching and to maintain the slew rate substantially constant.

Next, another example will be described, which has a structure for preventing capacitance as a load from changing even if the gate capacitance changes, instead of adjusting the drive current in accordance with the change of the gate capacitance, for the same purpose as the second embodiment.

Third Embodiment

Under the condition of the same drive current (drive capacity) of the pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b</i>, the rise time Tr and the fall time Tf of the delivered signal waveform becomes longer as capacitance of the load (gate capacitance) is larger, while the rise time Tr and the fall time Tf become shorter as the capacitance of the load is smaller.

Therefore, if the drain and the gate of the MOSFET's of the driver circuits <b>12</b><i>a </i>and <b>12</b><i>b </i>are connected by capacitance Cm, an effect of increasing the gate capacitance by Cm×gm×RL is obtained as a mirror effect. Here, gm is transconductance between the MOSFET's of the driver circuits <b>12</b><i>a </i>and <b>12</b><i>b</i>, and RL is load resistance.

If the number of MOSFET's that are turned on in the driver circuits <b>12</b><i>a </i>and <b>12</b><i>b </i>increases or decreases, gm changes in proportion to the number. Therefore, true gate capacitance and mirror capacitance also increase or decrease in proportion to it. Therefore, in order to prevent variations of the rise time Tr and the fall time Tf that may vary as they are, a total gate capacitance value is controlled to be constant by connecting or disconnecting appropriate capacitance for compensation between the drain and the gate at the same time as switching of the MOSFET.

More specifically, if the number of MOSFET's of the driver circuits <b>12</b><i>a </i>and <b>12</b><i>b </i>is large (if the gate capacitance is large), the capacitance for compensation is decreased. On the contrary, if the number of MOSFET's is small, the capacitance for compensation is increased. A third embodiment described below is one of examples of the method.

In the third embodiment, fundamental portions are the same as those of the first and second embodiments, so only different portions will be described.

<figref idrefs="DRAWINGS">FIG. 7</figref> is a circuit diagram showing a driving portion <b>3</b>C of an interface circuit <b>1</b>C according to the third embodiment of the present invention.

In <figref idrefs="DRAWINGS">FIG. 7</figref>, the driving portion <b>3</b>C is made up of a pre-buffer circuit <b>11</b>, a driver circuit <b>12</b>, a terminating resistor <b>13</b> and a capacitance switching circuit <b>18</b>. The capacitance switching circuit <b>18</b> includes two capacitance switching circuits <b>18</b><i>a </i>and <b>18</b><i>b. </i>

As understood from comparison between <figref idrefs="DRAWINGS">FIGS. 7 and 2</figref>, the capacitance switching circuit <b>18</b><i>a </i>is inserted as a load of one of the pre-buffers <b>11</b><i>a </i>in the driving portion <b>3</b>C, while the capacitance switching circuit <b>18</b><i>b </i>is inserted as a load of one of the other pre-buffer <b>11</b><i>b. </i>

The capacitance switching circuit <b>18</b><i>a </i>includes four capacitors <b>181</b>-<b>184</b> and four switches <b>191</b>-<b>194</b>. When the switches <b>191</b>-<b>194</b> are turned on, the corresponding capacitors <b>181</b>-<b>184</b> are connected as a load of the pre-buffer <b>11</b><i>a. </i>

The four capacitors <b>181</b>-<b>184</b> have the same capacitance as the four MOSFET's <b>121</b>-<b>124</b>. The switches <b>191</b>-<b>194</b> are controlled to be turned on and off by inverted signals of the control signals D<b>5</b>, D<b>6</b>, D<b>3</b> and D<b>4</b>, respectively. For example, if the control signal D<b>5</b> is turned on (“H”), the inverted signal thereof becomes off (“L”), and the corresponding switch <b>191</b> is turned off.

The capacitance switching circuit <b>18</b><i>b </i>also includes four capacitors <b>186</b>-<b>189</b> and four switches <b>196</b>-<b>199</b> in the same manner. When the switches <b>196</b>-<b>199</b> are turned on, the corresponding capacitors <b>186</b>-<b>189</b> are connected as a load of the pre-buffer <b>11</b><i>b. </i>

The four capacitors <b>186</b>-<b>189</b> have the same capacitance as those of the four MOSFET's <b>126</b>-<b>129</b>, respectively. The switches <b>196</b>-<b>199</b> are controlled to be turned on and off by inverted signals of the control signals D<b>1</b>-D<b>4</b>, respectively. For example, when the control signal D<b>1</b> is turned on (“H”), the inverted signal thereof becomes off (“L”), and the corresponding switch <b>196</b> is turned off. Note that NOT circuits are used, for example, for obtaining the inverted signals of the control signals D<b>1</b>-D<b>6</b>.

Therefore, even if one of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> is connected as a load of the pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b</i>, the capacitors <b>181</b>-<b>184</b> and <b>186</b>-<b>189</b> are connected having the same capacitance value as those of the MOSFET's that are not connected. Thus, a value of the capacitance to be a load of the pre-buffers <b>11</b><i>a </i>and <b>11</b><i>b </i>is always constant.

In this way, according to the interface circuit <b>1</b>C of the third embodiment, even if process variation or temperature variation occurs, it is possible to realize impedance matching and to maintain the slew rate substantially constant.

Note that it is possible in the third embodiment to set the capacitance values of the capacitors <b>181</b>-<b>184</b> and <b>186</b>-<b>189</b> to be independent capacitance values without associated with the gate capacitance of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b> and to generate the control signal for controlling the switches <b>196</b>-<b>199</b> by an appropriate logic instead of being the inverted signals of the control signals D<b>1</b>-D<b>6</b> so that the capacitance can be compensated in accordance with the switching of the MOSFET's <b>121</b>-<b>124</b> and <b>126</b>-<b>129</b>. In this case, the capacitance values of the capacitors <b>181</b>-<b>184</b> and <b>186</b>-<b>189</b> and the control signal for controlling the switches <b>196</b>-<b>199</b> should be decided so that an optimal result can be obtained by a simulation or the like, for example.

Fourth Embodiment

Next, another example of the calibration circuit will be described as a fourth embodiment.

<figref idrefs="DRAWINGS">FIG. 8</figref> is a circuit diagram showing a calibration portion <b>4</b>D of the fourth embodiment.

In a calibration circuit <b>14</b>D that constitutes the calibration portion <b>4</b>D, voltages VC<b>1</b>-VC<b>3</b> generated by constant current from the constant current sources <b>31</b>, <b>41</b> and <b>52</b> that flows in a MOSFET <b>32</b>, a resistor <b>42</b> and a MOSFET <b>51</b>, respectively, are selected one by one by a selector <b>22</b> and is supplied to an A/D convertor <b>23</b>. The A/D convertor <b>23</b> converts the analog voltages VC<b>1</b>-VC<b>3</b> into digital data VCD<b>1</b>-VCD<b>3</b> having eight bits or other number of bits with high accuracy and delivers it to an operational control portion <b>24</b>. The operational control portion <b>24</b> performs computing based on the values of the digital data VCD<b>1</b>-VCD<b>3</b> and generates the control signals D<b>1</b>-D<b>6</b>. The switches <b>131</b>-<b>139</b> and the like are controlled based on the generated control signals D<b>1</b>-D<b>6</b>. Note that the operational control portion <b>24</b> controls the selector <b>22</b> based on the signal SS so that the voltages VC<b>1</b>-VC<b>3</b> are supplied selectively, and it fetches the digital data VCD<b>1</b>-VCD<b>3</b> at appropriate timing.

Note that computing contents and process contents in the operational control portion <b>24</b> should be decided by a simulation or the like so that the process variation and temperature variation can be detected appropriately and that impedance matching can be performed appropriately based on the generated control signals D<b>1</b>-D<b>6</b>.

In this way, if the calibration circuit <b>14</b>D is used, the voltages VC<b>1</b>-VC<b>3</b> generated in the MOSFET <b>32</b>, the resistor <b>42</b> and the MOSFET <b>51</b> are detected by the A/D convertor with high accuracy, and control contents can be set delicately by computing in the operational control portion <b>24</b>. Therefore, according to the calibration portion <b>4</b>D of the fourth embodiment, it is possible to perform impedance matching and stabilizing of the slew rate with higher accuracy and more precisely than the case where the calibration circuit <b>14</b> of the first embodiment is used. Note that the calibration circuit <b>14</b> of the first embodiment has an advantage of a simple circuit and a small mounting area.

In the embodiments described above, the MOSFET's <b>120</b> and <b>125</b> correspond to the commonly used transistors in the present invention. The driver circuit <b>12</b> corresponds to the output circuit of the semiconductor integrated circuit in the present invention. The slope switching circuit <b>15</b> corresponds to the current switching circuit in the present invention. The constant current sources <b>151</b> and <b>154</b> correspond to the second constant current circuits in the present invention. The capacitance switching circuit <b>18</b> corresponds to the capacitance circuit in the present invention. In addition, the above-mentioned LSI in which the interface circuit <b>1</b> or <b>1</b>B-<b>1</b>D corresponds to the semiconductor integrated circuit in the present invention.

In the embodiments described above, it is possible to provide various necessary circuits after the terminating resistor <b>13</b>. In addition, the pre-buffer circuit <b>11</b>, the driver circuit <b>12</b>, the configuration of circuit, the structure thereof, the element constant, the number of elements and the like of a part or a whole of the terminating resistor <b>13</b>, the calibration circuit <b>14</b> or <b>14</b>D, the slope switching circuits <b>15</b> and <b>18</b> and the interface circuit <b>1</b> or <b>1</b>B-<b>1</b>D can be modified variously in accordance with the spirit of the present invention.

Although the embodiments of the present invention are described above, the present invention can be embodied variously without limiting to the above embodiments.